Limitations of the Current Child Sexual Abuse Medical Literature from a Bayesian Perspective

Abstract

This article explores whether the current child sexual abuse literature supports the use of Bayes's Theorem as a means of incorporating medical examination findings into child protection decisions. The purpose of the article is to determine whether likelihood ratios for specific physical findings are available from the current medical literature and to describe deficiencies in the quality of the quantitative medical information available to child abuse decision makers. The article reviews the structure of the current literature from a medical decision-making framework. Of the articles providing quantitative estimates of the value of different physical findings for decision making, only one article allows for the calculation of likelihood ratios for physical examination findings in prepubertal females in a form that facilitates decision making. This article fails to address the interaction between different medical findings. In addition, the examiners in this study were not blinded to whether patients examined had been sexually abused. Although the current child abuse literature can provide estimates of the strength of different physical examination findings, decision makers should be aware of the limitations of these estimates.